Photosensitizers Neutral Red (Type I) and Rose Bengal (Type II

Marks, G. T.; Lee, E. D.; Aikens, D. A.; Richtol, H. H. Photochem. ...... Tianlie Luo , Jingwen Chen , Bo Song , Hua Ma , Zhiqiang Fu , Willie J.G.M. ...
0 downloads 0 Views 292KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 6307-6313

Photosensitizers Neutral Red (Type I) and Rose Bengal (Type II) Cause Light-Dependent Toxicity in Chlamydomonas reinhardtii and Induce the Gpxh Gene via Increased Singlet Oxygen Formation BEAT B. FISCHER,† ANJA KRIEGER-LISZKAY,‡ AND R I K I . L . E G G E N * ,† Department of Environmental Microbiology and Molecular Ecotoxicology, Swiss Federal Institute for Environmental Science and Technology (EAWAG), Ueberlandstrasse 133, 8600 Du ¨ bendorf, Switzerland, and Institute of Biology II, University of Freiburg, Scha¨nzlestrasse 1, 79104 Freiburg, Germany

The connection between the mode of toxic action and the genetic response caused by the type I photosensitizer and photosynthesis inhibitor neutral red (NR) and the type II photosensitizer rose bengal (RB) was investigated in the green alga Chlamydomonas reinhardtii. For both photosensitizers, a light intensity-dependent increase in toxicity and expression of the glutathione peroxidase homologous gene (Gpxh) was found. The toxicity of RB was reduced by the singlet oxygen (1O2) quenchers 1,4diazabicyclo[2.2.2]octane and L-histidine, and the RB-induced Gpxh expression was stimulated in deuterium oxidesupplemented growth medium. These observations clearly indicate the involvement of 1O2 in both toxicity and the genetic response caused by RB. NR up-regulated the expression of typical oxidative and general stress response genes, probably by a type I mechanism, and also strongly induced the Gpxh expression. The stimulating effect of deuterium oxide in the growth medium suggested the involvement of 1O2 also in the NR-induced response. Indeed, an increased 1O2 formation was detected with EPR-spin trapping in NR-treated spinach thylakoids. However, none of the 1O2 quenchers could reduce the light-dependent toxicity of NR in C. reinhardtii, indicating that NR has a different mode of toxic action than RB.

Introduction Sunlight is essential as energy source for photosynthetic organisms, but it can also harm the cell (1, 2). By absorbing visible light, exogenous or endogenous photosensitizers increase the probability of light-induced damages. Once excited, photosensitizers become highly reactive and interact with cellular components disturbing their function (3, 4). Chloroplasts contain many potential photosensitizers, especially the chlorophylls, that are linked with chlorophyll* Corresponding author phone: +41-(0)1-823 5320; fax: +41(0)1-823 5547; e-mail: [email protected]. † EAWAG. ‡ Universita ¨ t Freiburg. 10.1021/es049673y CCC: $27.50 Published on Web 08/28/2004

 2004 American Chemical Society

binding proteins to reduce their photosensitizing properties. Photosynthetic organisms have optimized defense mechanisms (e.g., efficient 1O2 quenchers such as carotenoids and R-tocopherol) to protect themselves from light-induced damages under normal conditions (1, 5, 6). Under extreme situations, however, such as illumination with high light intensities or inhibition of the photosynthetic electron transport, the capacity of the defense mechanisms may be too limited (7-10). As a consequence, cells encounter an oxidative stress condition with subsequent damage to cell components. Beside these natural photosensitizers, pollutants, pesticides, and their degradation products have been shown to act as photosensitizers and to be toxic to organisms in the presence of light (11, 12). Two types of reaction mechanisms are involved in photosensitizer-caused toxicity: In type I reactions, the excited photosensitizer reacts with a substrate by direct electron transfer, resulting in a semi-reduced radical form of the substrate (13). Type II reactions involve the formation of singlet oxygen (1O2) via energy transfer from the excited photosensitizer to ground-state molecular oxygen (3O2) (14). Typical examples of types I and II photosensitizers are neutral red (NR) and rose bengal (RB), respectively. RB is known to mainly generate 1O2 by absorbing light in the visible range, and it is often used to study 1O2-mediated responses in vivo (15-18). Negative effects of RB on the photosynthetic electron transport chain in photosystems I and II (PSI and PSII) have been studied before (19, 20). NR, a phenazine-based dye, is widely used for staining cellular particles and as an intracellular pH indicator (21, 22) and was reported to directly transfer electrons in the excited state to several substrates in a type I manner (13, 21, 23). Additionally, NR accumulates in the thylakoid lumen of photosynthetic organisms upon illumination with concomitant changes in the chloroplast ultrastructure (24). As a result, a leak of protons from the thylakoid lumen and an inhibition of the electron transport chain in isolated chloroplasts were observed, and different pH-dependent inhibition mechanisms were discussed (25). NR therefore may have multiple effects on photosynthetic organisms, which may result in various cellular responses. Recently, we have investigated the genetic response of the green alga Chlamydomonas reinhardtii to the type I photosensitizer NR and the type II photosensitizer RB with DNA microarrays and found many general and oxidative stress genes to be induced by NR, but only one gene, the glutathione peroxidase homologous gene Gpxh, was strongly induced by both photosensitizers (unpublished data). Gpxh was shown to be specifically induced by the action of photosensitizers with very similar kinetics, indicating that a concerted signal cascade may be involved in the activation of Gpxh by the different photosensitizers (26, 27). The production of 1O2 was hypothesized to be a key intermediate in the response of the Gpxh expression to both NR and RB, even though NR normally reacts as type I photosensitizer. However, we did not have a direct proof for the involvement of 1O2 in the Gpxh induction, nor could we link the toxic action of these photosensitizers with 1O2 or with genetic responses. Knowing these links would allow prediction of the toxic potential of a chemical from the genetic response induced already at low concentrations, an approach that is becoming more important in ecotoxicology. This knowledge builds the bases for the interpretation of data obtained by toxicogenomics and for the development of molecular biomarkers or bioassays using reporter genes (28). But before such biomarkers can be applied for the hazard assessment of unknown chemicals or environmental samples, the VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6307

responsible mechanisms for their induction must be carefully studied (29). In this work, we used RB and NR to study the toxic effects of photosensitizers, the responsible mechanisms, and the links with genetic responses. The type II photosensitizer RB caused a light intensity-dependent toxic effect and induction of the Gpxh expression. Both were linked to the increased formation of 1O2. NR, on the other hand, provoked a type I photooxidative stress response, an increased formation of 1O , and interrupted photosynthesis. As a result, NR caused 2 a light-dependent toxic action and multiple genetic stress responses.

Experimental Section Strains and Culture Conditions. C. reinhardtii strain cw15arg7mt- (CC-1618) and strain cw15arg7mt- containing the plasmid pASPro1 (26) were inoculated in Tris-acetatephosphate (TAP) medium (30) or TAP medium containing 50% deuterium oxide (D2O) and agitated on a rotatory shaker (150 rpm) under constant illumination with white light of 150 µmol m-2 s-1 (except otherwise mentioned) at 25 °C. All media were supplemented with 50 mg/L ampicillin and 50 mg/L arginine when required. For chlorophyll fluorescence measurements, cells were harvested during exponential growth phase, resuspended in high salt minimal medium (HSM) (31), and further incubated until they reached a density of 6 × 106 cells/mL. Growth Experiments, Stress Treatment, and Arylsulfatase Assay. An overnight culture of strain cw15arg7 mtpASPro1 was grown to a cell density of 6 × 106 cells/mL. Then, cells were washed by centrifugation at 1000 rpm for 10 min and re-suspended in fresh TAP medium to a cell density of 6 × 106 cells/mL. Aliquots of 5 mL of the culture were distributed in 6-well culture plates, and the appropriate amount of each chemical was added. Cell density was analyzed 0, 1, 3, and 5 h after incubation by measuring the optical density at 750 nm in a 1:5 dilution. An average growth rate over 5 h was calculated for each culture out of four timepoint measurements. A decrease in optical density was expressed as negative growth rate and expected to describe a lethal effect of the treatment. After 2, 4, and 6 h incubation, a 300 µL sample was taken, and the accumulation of the arylsulfatase in the medium was analyzed by an arylsulfatase assay as described (26). The average reporter gene induction over the 6 h was calculated by dividing the arylsulfatase activity of the stressed culture by the activity of the control culture. For experiments in D2O-containing medium, cells were adapted to the medium for 3-5 d by serial dilution in fresh medium to ensure optimal growth (32). Subsequently, a 10 mL culture in a 100 mL Erlenmeyer flask was incubated with the appropriate concentration of each chemical for 60 min before samples were taken. RNA Isolation and Real Time RT-PCR. For RNA isolation, cells of 10-20 mL of individual cultures were harvested by centrifugation, and total RNA was isolated by the acid guanidine isothiocyanate-phenol-chloroform method (33) using TRIzol Reagent (Life Technologies Ltd.) following the suppliers instructions. Sample concentration was adjusted to 3 µg/µL total RNA, and RNA quality was checked by agarose gel electrophoreses and ethidium bromide staining. For reverse transcription, 1 µg of DNase I-treated total RNA was incubated in a 50 µL reaction including 10 µL of a 5× reaction buffer (Invitrogen), 2 µL of oligo(dT)18 primer (0.1 mM), 5 µL of dNTP (5 mM), 5 µL of 0.1 M dithiothreitol (DTT), and 300 units of SuperScript reverse transcriptase (Invitrogen) for 1 h at 37 °C. The reaction was stopped by heating at 95 °C for 5 min, and the volume was adjusted to a concentration of 20 ng/µL of original RNA quantity for each sample. 6308

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004

Sequences of the primers for real time RT-PCR were designed with the Primer Express software (Applied Biosystems) using the sequence of the 3′ untranslated region of each gene as a template. Real time RT-PCR reactions were performed on the ABI Prism 7000 Sequence Detection System (Applied Biosystems) using the SYBR Green PCR Master Mix kit as recommended by the manufacturer. Primer concentrations were optimized for each gene. To evaluate unspecific amplification by primer dimers or contaminations, annealing profiles of PCR products were analyzed and control reactions without cDNA were performed. Threshold cycle (Ct) values were determined for all reactions in the logarithmic amplification phase, and the average Ct value was calculated for each sample out of three replicates. The Ct value of the gene coding for the Rubisco small subunit (Rbcs2) was used for normalization. Induction factors for the different conditions were calculated for each gene as suggested by the manufacturer, in three independent experiments. Preparation of Thylakoid Membranes. Chloroplasts were prepared according to Laasch (34) from market spinach. Chloroplasts were osmotically shocked prior to measurements of photosynthetic electron transport or spin-trapping EPR. Measurements were performed in a medium containing 0.3 M sucrose, 50 mM KCl, 1 mM MgCl2 and 25 mM Hepes (pH 7.6). The chlorophyll content was set to 20 µg of Chl/mL. Oxygen Evolution and Chlorophyll Quantification. Photosynthetic O2 evolution in isolated spinach thylakoids and intact C. reinhardtii cells was measured with a Clarktype oxygen electrode (Hansatech) using either a 20 µg of Chl/mL of thylakoid membrane suspension illuminated with saturating light or a 12 µg of Chl/mL of C. reinhardtii culture at 25 °C at 200 µmol m-2 s-1 light. With spinach thylakoids, 100 µM methyl viologen was added as electron acceptor and 5 mM ammonium chloride was added as uncoupler. The chlorophyll content was quantified according to Arnon (35). Chlorophyll Fluorescence Measurements. Chlorophyll fluorescence was measured with a pulse-amplitude modulated fluorometer (PAM) 101 (Walz, Germany) (36). The F0level was monitored by weak modulated measuring light (about 0.05 µmol m-2 s-1), and fluorescence rise was induced by a 50 ms pulse of saturating white light (8000 µmol m-2 s-1) using a xenon flash XMT 103. Before measuring, cells were exposed to the appropriate treatment and incubated for 15 min in the dark at 25 °C with continuous stirring. Spin-Trapping of 1O2 by TEMP. Spin-trapping assays were performed with 10 mM 2,2,6,6-tetramethylpiperidine (TEMP) and 30 mM methanol. Samples were illuminated for 10 min with different light intensities at 20 °C. X-band EPR spectra were recorded in a flat cell with a Bruker ESP 300 spectrometer at room temperature with 9.7 GHz microwave frequency, modulation frequency 100 kHz, and modulation amplitude 2 G. The spin trap was purified as described by Fufezan et al. (37). Relative 1O2 levels were calculated as the difference between the signals of the treated and the untreated thylakoid sample or between the treated sample and the background (same sample, not subjected to illumination), respectively.

Results Toxicity of NR and RB under Different Light Intensities. In contrast to normal toxic substances, the toxicity of photosensitizers is rather complex in that it increases with both concentration and light intensity (16, 38, 39). Therefore, we first studied the effect of increasing concentrations of NR and RB on growth of C. reinhardtii under two different light intensities and compared the results to the effect caused by the PSII inhibiting herbicide Dinoterb (DT). Cells of strain cw15arg7mt- pASPro1 were exposed to light intensities of 100 or 200 µmol m-2 s-1 for 6 h, and the average growth rate was calculated for each condition by measuring the cell densities. A concentration-dependent decrease of growth rate could

FIGURE 1. Effect of different light intensities and increasing concentrations of the photosensitizers RB (squares) and NR (diamonds) or the herbicide DT (triangles) on (a) the average growth rate or (b) the average induction of a Gpxh-arylsulfatase reporter gene construct was measured in C. reinhardtii cultures over 6 h. The cells were exposed to either light intensities of 100 (closed symbols) or 200 (open symbols) µmol m-2 s-1 at 25 °C. Expression of the reporter gene construct was normalized for average cell numbers, and induction factors as compared to the untreated culture at 100 µmol m-2 s-1 light were calculated. Average induction factors and standard errors of three independent measurements are shown. be measured for both photosensitizers, reaching the level of no growth in the lower light intensity at about 4 µM for NR and between 0.5 and 0.6 µM for RB, indicating that these may be the lethal concentrations (Figure 1a). At higher light intensities (200 µmol m-2 s-1), this border was already between 1.5 and 2.0 µM for NR and between 0.3 and 0.4 µM for RB. On the contrary, DT did not result in a significant lower lethal concentration at the higher light intensity (Figure 1a). To know whether the same mechanisms could underlie both the toxicity and the Gpxh induction by RB and NR in C. reinhardtii, we compared the induction of the Gpxh gene by RB, NR, and DT at the two different light intensities (Figure 1b). For this experiment, the expression of a reporter gene construct integrated into the genome, containing a fusion of the Gpxh promoter to the arylsulfatase gene, was analyzed by measuring the enzyme activity of the arylsulfatase after 2, 4, and 6 h (26). With these activities, we then calculated an average induction of the Gpxh reporter gene over the 6 h exposure. The Gpxh expression was induced by all three substances with a concentration-dependent increase up to the level where the toxic effect of the substances dramatically reduced the growth rate and thus presumably the cell viability (Figure 1b). Up to these cytotoxic concentrations, the reporter gene expression was strongly stimulated by the elevated light intensity with both photosensitizers. Additionally, there was already a 1.5-fold increase in the ground-state expression at the higher light intensity as compared to the lower light intensity, indicating that there is some induction by endogenous photosensitizers under this condition. Even though we could measure an induction by the herbicide DT, there was no significant difference between the expressions under the two light conditions. Effect of 1O2 Quenchers. Formation of 1O2 is one possible toxic effect caused by photosensitizers to organisms, and

FIGURE 2. Average growth rate (open symbols) and the average induction of the Gpxh-arylsulfatase reporter gene construct (closed symbols) over 6 h affected by increasing concentrations of the photosensitizers RB (a) and NR (b) at 100 µmol m-2 s-1 illumination without quencher (triangles) or in the presence of 8 mM DABCO (squares) or 20 mM L-His (diamonds). the addition of exogenous 1O2 quenchers may have a protective effect on the cells exposed to type II photooxidative stress (40-42). To test, whether higher rates of 1O2 production are causing the cell death upon exposure to RB and NR, two widely used quenchers of 1O2, 1,4-diazabicyclo[2.2.2]octane (DABCO) and L-histidine (L-His), were added to cultures of the C. reinhardtii strain cw15arg7 mt- pASPro1 exposed to increasing concentrations of RB or NR at 100 µmol m-2 s-1 light. Even though L-His can be used by C. reinhardtii as a nitrogen source (43), its uptake is probably limited under the experimental conditions used (44). The effect of these quenchers on toxicity by the photosensitizers was measured by following growth for 6 h. Without quenchers, the average growth rate decreased dramatically above 0.5 µM RB and 3 µM NR (Figure 2). Upon addition of the 1O2 quenchers DABCO (8 mM) or L-His (20 mM), no decreased growth rate was measured up to 1 µM RB (Figure 2a). The NR-caused cytotoxicity could not be neutralized by the addition of the two quenchers. Surprisingly, even an increased toxicity of NR was detected in the presence of 8 mM DABCO (Figure 2b). To detect possible effects of 1O2 quenching on the induction of the Gpxh gene by RB and NR, the expression of the Gpxh-arylsulfatase reporter gene was measured during the 6 h of exposure. No significant difference in the Gpxh expression between quencher-containing and control culture was measured at low RB and NR concentrations, where toxicity does not affect the expression of the enzyme (Figure 2). Since DABCO has a negative effect on the growth of NRexposed cells, the induction of Gpxh is reduced already at lower concentrations as compared to the control, whereas no difference was seen for L-His. Addition of the quenchers to RB-exposed cells reduced toxicity of the photosensitizer and positively influenced the Gpxh expression at higher RB levels (Figure 2a). Consequently, the induction of Gpxh further increased or remained high at concentrations above 0.5 µM RB reaching a maximum level at 0.7-1 µM. No direct effect of DABCO on the Gpxh expression could be observed at low photosensitizer concentrations, and L-His only slightly reduced the Gpxh induction between 0.3 and 0.5 µM RB but not at concentrations below 0.2 µM RB (Figure 2a). Induction Profiles by RB and NR in D2O-Supplemented Growth Medium. Another often used method to test the VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6309

TABLE 1. Inhibition of Photosynthetic O2 Evolution by NR spinach thylakoids

FIGURE 3. Expression profiles of four genes in C. reinhardtii cells after exposure to 0.3 µM RB, 1 µM NR, or 2 mM H2O2 for 60 min in either a normal (gray bars) or a 50% D2O-containing (black bars) medium. Each induction factor was calculated as an average of three independent experiments with standard error bars. (a) Gpxh, (b) GST (BE129393), (c) HSP (BF864288), (d) Tub2B. involvement of 1O2 in photosensitized processes is the exchange of water (H2O) by deuterium oxide (D2O) in the reaction medium. The presence of D2O increases the lifetime of 1O2 in aqueous solutions and subsequently increases the frequency of 1O2-caused substrate modifications (40, 45, 46). We studied the effect of D2O-supplemented growth medium on the expression of several stress genes, known to be induced by RB and/or NR and thus analyzed, whether 1O2 may be involved in the response of these genes upon exposure to these photosensitizers. First, a culture of strain cw15arg7 mtwas pregrown for 3-4 d in TAP medium containing 50% D2O to adapt the cells and allow the cells to overcome the D2Ocaused stress (47, 48). Subsequently, the cells were exposed to a low concentration of the photosensitizer RB (0.3 µM) or NR (1 µM) and to 2 mM H2O2, inducing an oxidative stress not involving 1O2. Total RNA was isolated after 60 min exposure, and the mRNAs of the stress response genes Gpxh, BE129393, and BF864288 and the structural gene Tub2B were quantified with real time RT-PCR and normalized using the Rbcs2 gene as a reference. Induction factors for each gene were calculated as compared to the unexposed control culture in normal TAP medium. As shown earlier, the Gpxh gene was induced by both NR and RB even at low concentrations, whereas only little response was measured upon H2O2 exposure (Figure 3a) (26, 27). The induction of Gpxh by both RB and NR was 2-fold higher in the D2O-containing medium, indicating that indeed 1O , with a longer lifetime in D O, is responsible for the Gpxh 2 2 induction by the two photosensitizers. The response of the Gpxh expression to H2O2 was not or only to a minor extent 6310

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004

NR

photosynth acty µmol of O2/ (mg of Chl*h)

control 10 µM 20 µM 40 µM

422 399 312 234

C. reinhardtii

inhibition (%)

photosynth acty µmol of O2/ (mg of Chl*h)

inhibition (%)

5 26 45

24 23.8 21.3 14.8

1 11 38

influenced by the D2O-containing medium. Remarkably, the expression of Gpxh in the untreated control was already higher in the D2O-containing medium as compared to normal TAP medium. This could be either due to a higher 1O2 production under the D2O-dependent stress condition or, more likely, due to the stabilization of 1O2, which is produced in normal photosynthesis and presumably involved in the basal expression of Gpxh in the light. Additionally, the expression of two genes that we had identified in a previous study as being strongly induced by NR but hardly by RB (unpublished data) was tested in D2Ocontaining medium. One gene, identical to EST BE129393 in the Chlamydomonas EST library (49) and most probably coding for a glutathione S-tranferase (GST), was strongly induced in the normal medium by NR and H2O2, indicating that it responds to several oxidative stress conditions (Figure 3b). H2O2-induced expression was not significantly altered by the presence of D2O, but for NR, and to a minor extent also for RB, there was rather a down-regulation by the D2Ocontaining medium as compared to the induction in normal medium. The expression of another stress gene, BF864288, homologous to a chloroplast heat shock protein (HSP) had a similar profile to that of the GST gene when exposed to RB, NR, and H2O2 in normal medium (Figure 3c). A slight upregulation was observed under all conditions in D2Ocontaining medium. As control, the expression of the Tub2B gene, coding for a β-tubulin gene, was tested under the various conditions. Tub2B expression hardly varied by the exposure to the various conditions, and the D2O-containing medium did not cause any significant effect (Figure 3d). In conclusion, these results clearly showed that NR induced the expression of genes both by the formation of 1O2 and in another oxidative stress-dependent manner whereas RB only caused a 1O2-mediated stress response. Inhibition of Photosynthetic Activity and Generation of 1O by NR in Isolated Thylakoids. NR has been shown to 2 inhibit the photosynthetic activity by different mechanisms (25). To check whether the induction of the Gpxh expression by NR may be due to an increased generation of 1O2 in the NR-inhibited photosynthetic apparatus as found for phenolic herbicides (37), inhibition of the photosynthetic activity by NR was tested. Therefore, the reduction of photosynthetic O2 evolution was measured both in isolated spinach thylakoid membranes and in intact C. reinhardtii cells. As already reported by Opanasenko et al., reduction of the electron transport by NR is maximally 50-60% (25). In agreement with this, we found an inhibition of photosynthetic activity of 10-45% between 20 and 40 µM NR both in the thylakoid suspension and in the C. reinhardtii culture (Table 1). No significant reduction was found at concentrations lower than 20 µM NR in C. reinhardtii; a 5% inhibition of linear electron transport was measured in the thylakoid membranes. Furthermore, chlorophyll fluorescence in dark-adapted C. reinhardtii cultures exposed to NR was analyzed and compared to the effects caused by RB (Table 2). Already in the presence of 1 µM NR the variable fluorescence parameter was reduced to 89% as compared to the untreated sample and further decreased to 76% of the normal level at 10 µM

TABLE 2. Effect of RB or NR Exposure on the Yield of the Variable Chlorophyll Fluorescence in C. reinhardtii NR

rel Fv/Fm

RB

rel Fv/Fm

1 µM 5 µM 10 µM

0.89 ( 0.05 0.81 ( 0.04 0.76 ( 0.05

0.5 µM 1 µM 2 µM

0.92 ( 0.03 0.97 ( 0.04 0.98 ( 0.05

NR. This indicates that a reduction of photosynthetic efficiency takes place already at low NR concentrations, which might be even more pronounced under illumination (25). RB, on the other hand, did not influence the variable chlorophyll fluorescence in the dark, indicating that no direct interaction of RB with the photosynthetic apparatus occurs (Table 2). Finally, the production of 1O2 in isolated thylakoid treated either with RB or NR was directly measured by EPR spintrapping (37). Therefore, thylakoids were exposed to increasing concentrations of RB or NR and illuminated with white light for 10 min at 150 or 1500 µmol m-2 s-1. As a type II photosensitizer, RB produced 1O2 both in an aqueous solution (not shown) and in the presence of thylakoids in a concentration-dependent manner (Figure 4a,b). In contrary, NR did not lead to the formation of any 1O2 in aqueous solutions up to 20 µM (see Figure 4a for 5 µM NR), but it significantly rose the production of 1O2 in a solution with isolated spinach thylakoids already at low concentrations as compared to an untreated thylakoid solution (Figure 4a,b). We additionally tested, whether the production of 1O2 is dependent on the light intensity. As expected, in the presence of the type II photosensitizer RB the 1O2 level strongly increased with higher light intensities already in the presence of 0.2 µM RB (Figure 4c). Similarly, NR also stimulated the production of 1O2 in a thylakoid suspension in a light-dependent manner with a significant increase as compared to the control already at light intensities between 100 and 200 µmol m-2 s-1 (Figure 4c).

Discussion Exposure of C. reinhardtii to exogenous photosensitizers caused severe damages to the cell resulting in cell death and the induction of genetic stress response. In this work we investigated the effects of different photosensitizers on C. reinhardtii, using the type I sensitizer NR and the type II sensitizer RB, to study the mode of toxic action and the intermediates involved in the toxicity. Furthermore, we linked the physiological responses caused by the two photosensitizers with genetic expression profiles and identified similarities and differences in the mechanisms of RB- and NRinduced cellular effects. RB-Induced Cellular Response. The toxicity and the Gpxh induction caused by RB in the C. reinhardtii cultures were both dependent on the light intensity and the concentration of the photosensitizer (Figure 1). The same characteristics were found for the RB-dependent 1O2 production, indicating that this photosensitizing property might cause the cellular effects (Figure 4). A role of 1O2 in these response is supported by the reduced toxicity in the presence of the 1O2 quenchers DABCO and L-His and the stimulated expression of Gpxh in the D2O-supplemented medium (Figures 2a and 3a). As control, another typical oxidative stress gene (GST) and a general stress response gene (HSP) that both strongly responded to oxidative stress caused by H2O2 and NR were not, or only to a minor extent, induced by RB in both the normal and the D2O-supplemented medium supporting the specific 1O2-dependent Gpxh induction (Figure 3b,c). A slightly higher induction of the HSP gene in the D2Ocontaining medium was found for all treatments tested. This is most probably due to an increased general stress caused

FIGURE 4. (a) Typical EPR spectra of TEMPO as adduct of the reaction of TEMP without 1O2 (top spectrum), with 1O2 produced by 5 µM NR in the absence (second spectrum) or the presence (third spectrum) of isolated spinach thylakoid membranes or with 1O2 produced by 0.5 µM RB (bottom spectrum) after 10 min illumination with 150 µmol m-2 s-1 PAR. (b) Concentration-dependent increase of 1O2 production in isolated thylakoids in the presence of RB (diamonds) or NR (triangles) illuminated with 1500 µmol m-2 s-1 PAR. (difference between treated and untreated thylakoid signal). (c) Light intensity-dependent increase of 1O2 production in isolated spinach thylakoids incubated either without any treatment (squares) or with 0.2 µM RB (diamonds) or 20 µM NR (triangles) for 10 min (difference between treated thylakoids and background signal). by the D2O-containing medium rather than due to any specific stress caused by 1O2 (47). Earlier expression analysis with DNA microarrays indicates that only a few genes are induced by 1O2 in C. reinhardtii, suggesting the involvement of a specific mechanism in the response to 1O2 (unpublished data). No obvious effect of the 1O2 quenchers DABCO and L-His on the expression of the Gpxh gene could be measured at VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6311

low RB concentrations, even though the reduced toxicity and concomitantly increased expression of the reporter gene construct at high concentrations indicate that 1O2 is significantly quenched by both substances (Figure 2a). However, the rate and site of production or removal of 1O2 by RB or DABCO and L-His, respectively, might not be homogeneous inside the cell (18). Local concentration of the photosensitizer in the cell was shown to be crucial for its mode of action and the efficiency earlier (38, 39). RB is taken up efficiently by the cell and is thought to localize near the lipid-aqueous interface with two carboxylic acid group exposed to the aqueous phase (50, 51). DABCO, on the other hand, was suggested to penetrate only to a low extent into the cell at physiological pH and therefore mainly protects against 1O2 damages in the cell membrane (52). Thus, it is possible that the sensing of 1O or its effect leading to the induction of the Gpxh expression 2 is locally different in the cell from the site where the toxic action of RB is taking place. Additionally, different target molecules of 1O2 have very different quenching efficiencies for 1O2 produced in the cell (51, 53). Therefore, 1O2-dependent activation of the cellular sensor responsible for the Gpxh expression might occur with a much higher efficiency than the modification of the toxic site. Thus, the quenching efficiency of DABCO and L-His for 1O2 may be too low to significantly reduce the activation of the sensor, whereas it may successfully compete with a lower quenching efficiency of the component responsible for the toxic action of 1O2. Different localization and/or quenching efficiency of the photosensitizer, the targets, and the quenchers may explain the converse effect of DABCO and L-His on the toxicity and the Gpxh expression caused by RB. NR-Induced Cellular Response. NR has been shown not only to act as a type I photosensitizer (13, 22, 23) but also to uncouple ATP synthesis from photosynthetic electron transport and to inhibit the electron transport chain (25). Even though the toxic effect of NR to the cell is unknown, the light-dependent reduction of growth rate indicates that a photosensitized process is responsible for the NR toxicity (Figure 1a). Therefore, two different mechanisms may be involved. In a type I mechanism, the excited NR could directly interact with cellular components in an electron-transfer reaction and thus damage essential cellular functions. This is supported by the strong induction of the oxidative stress gene GST and the HSP gene, which both were also induced by H2O2, indicating that a strong intracellular stress caused by a disturbed redox balance at low NR concentrations occurred (Figure 3b,c). The second possible mechanism for a light-dependent toxicity of NR could involve an increased formation of 1O2 in the chloroplast, as measured in NR-treated isolated spinach thylakoids by EPR spin-trapping, with an enhanced generation of 1O2 already at low NR concentrations (Figure 4). These levels of 1O2 produced in the thylakoid exposed to NR might still be underestimated when compared to 1O2 levels produced by RB in the medium because only 1O2 that diffused out of thylakoids is measured with the spin-trapping. Unfortunately, we cannot measure directly the production of 1O2 by NR in intact C. reinhardtii cells by chemical trapping techniques (data not shown). The induction of the Gpxh gene by NR and the stimulating effect of D2O-containing medium as compared to normal growth medium imply that 1O2 is indeed produced and that 1O2 mediates the concomitant response of the Gpxh gene (Figure 3a). Since the expression of the Gpxh gene increased linearly with the concentration of RB in the nontoxic range, we assume a proportional induction of Gpxh to the amount of 1O2 formed (Figures 1b and 2a). As a consequence the increased expression of Gpxh at 200 as compared to 100 µmol m-2 s-1 illumination at nontoxic NR concentrations (0.5 µM) (Figure 1b) is an indirect 6312

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004

evidence for a light intensity-dependent generation of 1O2 by NR in C. reinhardtii cells. Still, the mechanism of the increased 1O2 generation by NR in the thylakoids is unknown: NR might have a type II activity under certain conditions like, for example, low pH in the thylakoid lumen, and in addition, by inhibiting the photosynthetic electron flow, it might increase the production of 1O2 via charge recombination in PSII, as described for phenolic herbicides such as DT (7, 37). However, no significant reduction of the linear electron transport in NR-treated C. reinhardtii cultures or in isolated spinach thylakoids was measured with an O2 electrode at concentrations below 20 µM (Table 1). Similar results were also obtained by Opanasenko et al. (25) in isolated pea chloroplasts. The authors hypothesized that a NR-caused inhibition of photosynthetic O2 evolution may occur already at low concentrations but is compensated by a stimulating effect of NR on the electron transport chain by acting as an uncoupler similar to other amines. Indeed, an inhibitory effect of low NR concentrations was measured on the yield of variable chlorophyll fluorescence in dark adapted C. reinhardtii cell (Table 2). Interestingly, DABCO has also been shown to uncouple electron transport in thylakoids such as other amines (54). A synergistic uncoupling effect of NR and DABCO could therefore be responsible for the increased toxicity found by the combination of the two substances (Figure 2b). Similar to RB response, no direct effect on the Gpxh expression by NR could be measured by the 1O2 quencher DABCO and L-His, indicating that these quenchers have no access to the 1O2 molecules involved in the response (Figure 2b). This suggests that the sensor for Gpxh induction by 1O2 is located close to the source of 1O2 in NR-treated cell, probably in the thylakoids. Further support for the localization of the 1O2 sensor in the thylakoids is given by the similar induction strength achieved by the relatively low amount of 1O produced by NR in the thylakoids as compared to the 2 high yield of 1O2 produced by RB in the medium (Figures 1 and 4). So far, no sensor for 1O2 was characterized in the chloroplast, even though a strong genetic response to 1O2 was also found in Arabidopsis where a set of 70 genes was specifically induced in protochlorophyllide accumulating mutant exposed to light (55). In this paper, the authors speculate that the 1O2-dependent modification of lipids, resulting in the formation of lipid peroxides, might trigger the stress response similar to the generation of ceramide in mammalian cells causing the activation of the transcription factor AP-2 (56). We have strong evidence that NR causes several unrelated effects to C. reinhardtii cells, which result in a light-dependent toxicity and the strong induction of various stress response genes including the 1O2-dependent Gpxh induction. RB on the other side, only strongly induced the Gpxh gene by the formation of 1O2 and also its toxicity could be linked to 1O2 effects. This shows that the genetic response of an organism to a pollutant already at low concentrations can give a lot of important information about the specificity and severity of its harmful action, on the condition that the mechanisms responsible for the induction of the selected genes are wellcharacterized. Unraveling these mechanisms builds the basis for the development of molecular biosensors and for the interpretation of genetic expression data.

Acknowledgments This work was financially supported by Emhart Glass SA, Cham, Switzerland. We thank Katrin Brombacher, Karin Ru ¨ fenacht, and Alexander J. B. Zehnder for stimulating discussions and continuous support.

Literature Cited (1) Keren, N.; Ohad, I. In The Molecular Biology of Chloroplasts and Mitochondria in Chlamydomonas; Rochaix, J.-D., Michel, G.-

(2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31)

C., Sabeeha, M., Eds.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1998; Vol. 7, pp 569-596. Barber, J.; Andersson, B. Trends Biochem. Sci. 1992, 17, 61-66. Smith, K. C. The Science of Photobiology; Plenum Press: New York, 1981. Davies, M. J. Biochem. Biophys. Res. Commun. 2003, 305, 761770. Trebst, A. Z. Naturforsch., C: J. Biosci. 2003, 58, 609-620. Baroli, I.; Gutman, B. L.; Ledford, H. K.; Shin, J. W.; Chin, B. L.; Havaux, M.; Niyogi, K. K. J. Biol. Chem. 2004, 279, 6337-6344. Rutherford, A. W.; Krieger-Liszkay, A. Trends Biochem. Sci. 2001, 26, 648-653. Hideg, E.; Kalai, T.; Hideg, K.; Vass, I. Philos. Trans. R. Soc. London, Ser. B 2000, 355, 1511-1516. Aro, E. M.; Virgin, I.; Andersson, B. Biochim. Biophys. Acta 1993, 1143, 113-134. Barber, J. The Photosystems: Structure, Function, and Molecular Biology; Elsevier Science: Amsterdam/New York, 1992. Huang, X. D.; Zeiler, L. F.; Dixon, D. G.; Greenberg, B. M. Ecotoxicol. Environ. Saf. 1996, 35, 190-197. Joshi, P. C.; Misra, R. B. Biochem. Biophys. Res. Commun. 1986, 139, 79-84. Marks, G. T.; Lee, E. D.; Aikens, D. A.; Richtol, H. H. Photochem. Photobiol. 1984, 39, 323-328. Foote, C. S. Photochem. Photobiol. 1991, 54, 659-659. Allen, M. T.; Lynch, M.; Lagos, A.; Redmond, R. W.; Kochevar, I. E. Biochim. Biophys. Acta 1991, 1075, 42-49. Lambert, C. R.; Stiel, H.; Leupold, D.; Lynch, M. C.; Kochevar, I. E. Photochem. Photobiol. 1996, 63, 154-160. Stiel, H.; Teuchner, K.; Paul, A.; Leupold, D.; Kochevar, I. E. J. Photochem. Photobiol. B 1996, 33, 245-254. Kochevar, I. E.; Redmond, R. W. Methods Enzymol. 2000, 319, 20-28. Chung, S. K.; Jung, J. Photochem. Photobiol. 1995, 61, 383-389. Telfer, A.; Bishop, S. M.; Phillips, D.; Barber, J. J. Biol. Chem. 1994, 269, 13244-13253. Singh, M. K.; Pal, H.; Bhasikuttan, A. C.; Sapre, A. V. Photochem. Photobiol. 1999, 69, 529-535. Singh, M. K.; Pal, H.; Sapre, A. V. Photochem. Photobiol. 2000, 71, 44-52. Singh, M. K.; Pal, H.; Sapre, A. V. Photochem. Photobiol. 2000, 71, 300-306. Opanasenko, V. K.; Semenova, G. A.; Agafonov, A. V.; Gubanova, O. N. Biochem. Moscow 1996, 61, 1083-1087. Opanasenko, V.; Agafonov, A.; Demidova, R. Photosynth. Res. 2002, 72, 243-253. Leisinger, U.; Ru ¨ fenacht, K.; Fischer, B.; Pesaro, M.; Spengler, A.; Zehnder, A. J. B.; Eggen, R. I. L. Plant Mol. Biol. 2001, 46, 395-408. Leisinger, U.; Ru ¨ fenacht, K.; Zehnder, A. J. B.; Eggen, R. I. L. Plant Sci. 1999, 149, 139-149. Neumann, N. F.; Galvez, F. Biotechnol. Adv. 2002, 20, 391-419. Eggen, R. I.; Behra, R.; Burkhardt-Holm, P.; Escher, B. I.; Schweigert, N. Environ. Sci. Technol. 2004, 38, 58A-64A. Harris, E. H. The Chlamydomonas Sourcebook: A Comprehensive Guide to Biology and Laboratory Use; Academic Press: San Diego, 1989. Sueoka, N. Proc. Natl. Acad. Sci. U.S.A. 1960, 46, 83-91.

(32) Chorney, W.; Scully, N. J.; Crespi, H. L.; Katz, J. J. Biochim. Biophys. Acta 1960, 37, 280-287. (33) Chomczynski, P.; Sacchi, N. Anal. Biochem. 1987, 162, 156159. (34) Laasch, H. Planta 1987, 171, 220-226. (35) Arnon, D. I. Plant Physiol. 1949, 24, 1-15. (36) Schreiber, U.; Schliwa, U.; Bilger, W. Photosynth. Res. 1986, 10, 51-62. (37) Fufezan, C.; Rutherford, A. W.; Krieger-Liszkay, A. FEBS Lett. 2002, 532, 407-410. (38) Aveline, B. M.; Sattler, R. M.; Redmond, R. W. Photochem. Photobiol. 1998, 68, 51-62. (39) Aveline, B. M.; Redmond, R. W. Photochem. Photobiol. 1999, 69, 306-316. (40) de Mol, N. J.; Beijersbergen van Henegouwen, G. M.; Mohn, G. R.; Glickman, B. W.; van Kleef, P. M. Mutat. Res. 1981, 82, 2330. (41) Polo, L.; Presti, F.; Schindl, A.; Schindl, L.; Jori, G.; Bertoloni, G. Biochem. Biophys. Res. Commun. 1999, 257, 753-758. (42) Song, Y. Z.; An, J.; Jiang, L. Biochim. Biophys. Acta 1999, 1472, 307-313. (43) Hellio, C.; Veron, B.; Le Gal, Y. Plant Physiol. Biochem. 2004, 42, 257-264. (44) Hellio, C.; Le Gal, Y. Comp. Biochem. Physiol., Part A: Mol. Integr. Physiol. 1998, 119, 753-758. (45) Lee, P. C.; Rodgers, M. A. Photochem. Photobiol. 1987, 45, 7986. (46) Bilski, P.; Belanger, A. G.; Chignell, C. F. Free Radical Biol. Med. 2002, 33, 938-946. (47) Unno, K.; Kishido, T.; Morioka, M.; Okada, S.; Oku, N. Biol. Pharm. Bull. 2003, 26, 799-802. (48) Williams, A. J.; Morse, A. T.; Stuart, R. S. Can. J. Microbiol. 1966, 12, 1167-1173. (49) Shrager, J.; Hauser, C.; Chang, C. W.; Harris, E. H.; Davies, J.; McDermott, J.; Tamse, R.; Zhang, Z.; Grossman, A. R. Plant Physiol. 2003, 131, 401-408. (50) Lambert, C. R.; Kochevar, I. E. Photochem. Photobiol. 1997, 66, 15-25. (51) Kochevar, I. E.; Bouvier, J.; Lynch, M.; Lin, C. W. Biochim. Biophys. Acta 1994, 1196, 172-180. (52) Anderson, R. F.; Patel, K. B. Photochem. Photobiol. 1978, 28, 881-885. (53) Kochevar, I. E.; Lambert, C. R.; Lynch, M. C.; Tedesco, A. C. Biochim. Biophys. Acta 1996, 1280, 223-230. (54) Sabat, S. C.; Mohanty, P. Indian J. Biochem. Biophys. 1992, 29, 490-493. (55) op den Camp, R. G.; Przybyla, D.; Ochsenbein, C.; Laloi, C.; Kim, C.; Danon, A.; Wagner, D.; Hideg, E.; Gobel, C.; Feussner, I.; Nater, M.; Apel, K. Plant Cell 2003, 15, 2320-2332. (56) Grether-Beck, S.; Bonizzi, G.; Schmitt-Brenden, H.; Felsner, I.; Timmer, A.; Sies, H.; Johnson, J. P.; Piette, J.; Krutmann, J. EMBO J. 2000, 19, 5793-5800.

Received for review March 2, 2004. Revised manuscript received June 7, 2004. Accepted July 9, 2004. ES049673Y

VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6313